Move the wiki to CDO

[Rast]

Problem: It's hard to get a wiki login, and it's yet another password to remember.

Solution: Fork the wiki, host it on CDO, and use mantis for wiki logins.

[chequers]

Doing this without the agreement of the current wiki editors would be pointless. They would continue to edit the wiki in its original location and then there would be two copies of the wiki, and the "more official" looking one would be more out of date.

If you want to move the wiki, you need a plan which will bring the existing editors along with it, or can replace them with equivalent productivity.

FWIW I tried the former a year ago and the editors decided they would prefer to remain independent.

[Rast]

If you want to move the wiki, you need a plan which will bring the existing editors along with it, or can replace them with equivalent productivity.

My thought is that the Tavern folks would edit the wiki a lot if we were all automatically logged in as wiki editors simply by being logged into Mantis.

The chaosforge wiki isn't super active of late: http://crawl.chaosforge.org/index.php?t ... 00&days=30

[ion_frigate]

I've adminned on that wiki, and on NetHackWiki. I figure I can offer a few insights:

1) You absolutely need the agreement of the community there. Also, it would be better if, once the new wiki is migrated, you could get Kornel to take the old one down. I was with NetHackWiki when they moved from Wikia to NAO, due to the former's ever-increasing meddling in formatting and general commercialization of the platform. It took *months* for the new wiki to start showing up above the old one (which Wikia would not close down, choosing to call the migration a "fork") in Google searches, and that is a *major* source of wiki visits. This problem would be a million times worse if you still had some people claiming the chaosforge wiki is the "real" one. No one bothered to defend the old NetHack Wikia wiki, yet I *still* see people linking it on reddit from time to time, just because they don't know better.

2) For what it's worth, I support this idea. I can't speak for the community on the wiki now, of course. Maybe try an approach that's less "the current wiki is crap" and more "you'd get more active participation." Respect that people work hard on that wiki. It's not out-of-date because they're idiots. It's out of date because Crawl changes *really* fast, both in terms of metagame and actual game. These are volunteers who don't necessarily have 50+ hours to spare every six months - because believe me, that *is* how long it takes to update every last little change. It's not just a matter of iteratively going through the changelog. Imagine, all of a sudden, that the developers got rid of the partial torment protection granted by rN+. You wouldn't just have to change the articles on torment and negative energy. You'd also have to scan through the strategy sections for every monster and area that has torment, as well as all the items that grant rN+. Changes in the game spider out into multiple changes in the wiki like that, and it's a lot of work to maintain it. Respect that.

3) You'll also need to install MediaWiki on CDO, almost certainly. Whatever software is used for the current dev wiki is just awful by comparison. It's hard to edit with, and produces ugly-looking pages. Perhaps these are subjective considerations, but I think you'd find that the current community on the chaosforge wiki is likely to agree - people like to stay with software they're comfortable with. I can tell you also, the category system in MediaWiki is very helpful to editing (particularly around new releases) as well as navigation; as far as I can tell, the dev wiki doesn't support anything similar. Moreover, players expect game wikis to look like Wikipedia, and the software the dev wiki uses looks nothing like it. Lastly, going over to the dev wiki format would probably mean starting from scratch, as opposed to migrating what's currently on the chaosforge wiki. Say all you want about how inaccurate the chaosforge wiki is, starting over from nothing would be a much more intensive project, and could easily fizzle out.

4) Spam, spam spam spam lovely spam, wonderful spam. That's why it's so hard to register on the chaosforge wiki. I remember the dark days when spambots were attacking it every other day, and we didn't even have an active admin to block them - all we could do was revert. I have no idea why it gets hit with it so horribly - I never saw anything like that on NetHackWiki - but if the problem were to carry over to CDO, that could be a big problem. On a related note, we'd probably have to ask Kornel roughly how much traffic the wiki gets, and make sure that CDO can handle that. I'm guessing it can, but still.

[ZipZipskins]

Just gonna chime in and say that despite all the work that has been put into into nethackwiki, wikihack results still top google searches.

Source: I shamefully started playing nethack again recently for reasons unbeknownst to anyone, the lord included

So yeah, that shit is complex and difficult.

[Rast]

3) You'll also need to install MediaWiki on CDO, almost certainly. ... Moreover, players expect game wikis to look like Wikipedia, and the software the dev wiki uses looks nothing like it.

4) Spam, spam spam spam lovely spam, wonderful spam.

3. For sure. I never even considered the dev wiki as a possibility.

4. Would spam bots understand that they needed to register for Mantis before they could use the wiki? That alone might be the ultimate captcha. If not, mods and admins could ban them as they popped up.

[watertreatmentRL]

The problems people cite re: the wiki, e.g. out of date, bad opinions and advice, are symptoms of structural problems with wikis. You won't get around them with a change of URL.

The wiki was formerly linked by the main page. The suggested progression of "1.) linking to the wiki, 2.) not linking to the wiki, 3.) hosting the wiki" somehow doesn't add up.

[advil]

(Not speaking for the rest of the devteam, I haven't discussed this with anyone) I can see the appeal of this idea and I wish I thought it was possible, but the active devs+admins are already spread quite thin across fairly disparate server admin tasks, to the point where it's sometimes hard to even know who is most likely to be able to fix/update a particular thing. For example we didn't even manage to get CDO updated in time for the last release tournament, I'm not sure that someone installing & maintaining mediawiki is in the cards. I think the state of dokuwiki, mentioned above, and mantis for that matter, is symptomatic of this same problem -- they're both 7-8 years out of date, and for reasons can't be updated.

(There's also the issue that having the wiki be official in any sense would require some changes in dev consensus. Though it's incorrect to say that it isn't linked, it's right there on the main cdo page in "help & community".)

[watertreatmentRL]

It was linked from the old main page in 2014, then not in 2015. I guess it was readded when the site was redesigned.

[Rast]

The problems people cite re: the wiki, e.g. out of date, bad opinions and advice, are symptoms of structural problems with wikis.

Also symptomatic of having to send an email to even get permission to edit.

[ion_frigate]

The problems people cite re: the wiki, e.g. out of date, bad opinions and advice, are symptoms of structural problems with wikis. You won't get around them with a change of URL.

If that change of URL makes it easier for more people to edit the wiki, it absolutely can help. What you're citing don't strike me as symptomatic of structural problems with wikis - just symptomatic of a wiki with too few people doing too many things. Don't get me wrong, article ownership and "POV-pushing" are real problems with places like Wikipedia, but I've never seen them on either of the game wikis I've been on. Bad advice stays around because no one has had the time or inclination to remove it - don't forget, it doesn't just have to be removed, but also has to be replaced. This takes time, and having more editors means it's more likely to be the case that someone will take the time to rewrite the bad advice section.